The objective of this project is to further develop and validate a new outcome measure--the failure rate. By dissecting the death rate into its component parts, the adverse-occurrence rate and the failure rate, we hope to develop improved measures of hospital quality of care. The specific aims are: (1) To validate, using a much larger and more diverse data set, the findings of our earlier work which has shown that the probability of developing an in-hospital adverse-occurrence among patients undergoing cholecystectomy and prostatectomy is primarily a function of patient characteristics, while the probability of dying, given an adverse-occurrence (defined as the "failure rate") is more a function of hospital characteristics. Thus, we seek to confirm that the factors which predict failure are distinct from those that predict mortality or adverse-occurrence. (2) To examine the validity of the model for surgical procedures of higher technology intensity and complexity, and for medical conditions. This requires development of more refined specifications of the adverse-occurrence and failure. (3) Compare the rankings of hospitals using mortality rates, failure rates and adverse-occurrence rates over individual and grouped diagnoses and procedures. (4) Study in greater detail those hospitals with very disparate rankings between the death rate and the failure rate. Based on the above, the information value and content of these outcome measures will be assessed and recommendations as to their applications will be suggested.